No abstract

Due to sharp increase in damages from localized heavy rainfall-induced landslide disasters in Korea since 2000s, there has been increasing interest in researches on the development of effective, practical, and reliable decision-making supportive tools in the disaster management such as early warning systems and risk assessment frameworks. As one of representative outcomes from the researches, a citywide landslide early warning system (LEWS) was developed and implemented in the local government of Busan, the second largest city of Korea, and now under test-operation. With the brief overview on distinctive features of the system, the paper specifically focuses on the concept of debrisflow risk analysis results presented in real-time with the highest warning level, Emergency. Since the areas of Emergency can be theoretically interpreted as debris-flow source areas and they are to be progressively expanded according to the accumulated rainfall input data (e.g., continuous rainfall amount), the initial volume of debris-flow corresponding to a continuous rainfall amount was estimated, and accordingly, numerical simulations and quantitative analyses of debris-flow movements, vulnerabilities and socioeconomic properties of risk elements were conducted in the predicted deposition area. A case study was conducted for a vulnerable site to debris-flow in a mountain of Busan. Lastly, thresholds based on human vulnerabilities were introduced and discussed in order to supplement the limitations of risk information based on the building vulnerability.

No abstract